EUV Sources Continue to Fight the Fight

If extreme ultraviolet (EUV) lithography is ever going to reach high-volume production levels, the most critical development that has yet to be achieved is a high-power EUV source. This was much discussed at this year's SPIE Microlithography conference, as well as the International SEMATECH (ISMT) EUV Source Workshop, which took place immediately preceding the conference.

The most pressing technical issue for EUV lithography's success is the development of a source with a high repetition rate, sufficient brightness and lifetime, and sufficiently low off-band heating and particulate emissions. Output power — the key to high throughput — can be scaled by increasing repetition rate, conversion efficiency and collection efficiency. Increasing the repetition rate helps to increase power, but requires improved thermal management. And in a technology where current bragging rights means you can achieve a conversion efficiency of >1%, thermal management is no small matter. Increasing conversion efficiency, which also improves output power, can be achieved by optimizing plasma conditions, but many developers are also looking into new target materials with better efficiencies than the commonly used xenon target.

Source developers have come a long way in the past couple years, but they've clearly got a long way to go. Source parameters have improved by more than an order of magnitude since 2001, but they will have to improve by about another order of magnitude to meet production requirements. Requirements, as laid out by the toolmakers (Table), define source specifications at intermediate focus (IF), which is the illuminator entrance. Those requirements call for 115 W IF. Most source developers have yet to reach 10 W.

Developers have been exploring the potential of, primarily, laser-produced plasma (LPP) sources and gas discharge produced plasma (GDPP) sources. XTREME technologies GmbH — a joint venture between Lambda Physik and Jenoptik — is exploring both possibilities. The required lifetimes seem to be easier to achieve with LPP sources than with GDPP sources, according to Christian Ziener, but the LPP sources are substantially more complex and expensive to produce. Also, the ultimate power achievable with GDPP sources may be limited because of excessive heat load at the discharge components.

XTREME technologies' LPP source has a 10 kHz repetition rate, generating >2 W of IF output power. At the time Ziener presented his company's status in February, his group was awaiting delivery of a higher-power laser source that would raise average power from 500 W to ~1.3 kW.

The GDPP source is based on Z-pinch with sliding discharge pre-ionization. In continuous operation at 2 kHz repetition rate, it achieves IF power of 10 W. The company's XTS 13-35 has already been shipped to Exitech Ltd. for integration into its EUV microstepper. The source has been integrated with collector optics from Carl Zeiss, and the EUV tool is in process development at ISMT.

The company has been working mainly with xenon, but tin could achieve a conversion efficiency of 2% as opposed to xenon's 0.8%, in theory raising IF output power to 44 W.

Cymer Inc. has been developing a GDPP source it refers to as a dense plasma focus (DPF) discharge EUV source. It has a continuous repetition rate of 2 kHz, currently achieving an output power of 11.3 W IF, with near-future expectations of 45.4 W. Like other developers, Cymer is looking into tin targets, as well as LPP source designs.

Other companies working toward source solutions include Powerlase Ltd., which is working on an LPP source; Philips Extreme UV GmbH, a joint venture between Philips and Fraunhofer, whose hollow cathode triggered GDPP source has been integrated into an alpha tool at ASML; and JMAR Technologies, which is exploring various LPP source setups.

Japanese consortium Extreme Ultraviolet Lithography System Development Association (EUVA) is investigating an LPP source with a 10 kHz repetition rate and an average output power of >1 kHz. At 4 W IF now, researchers expect to reach 10 W IF by 2006. They plan for a commercial LPP-based EUV tool by around 2008. EUVA's GDPP source has a continuous repetition rate of 2 kHz, with an estimated output power at 2.7 W IF, and plans to reach >30 W by 2007.

As part of Europe's MEDEA+ initiative, Alcatel Vacuum Technology has been developing an EUV source along with Thales Laser SA and French national laboratory CEA Saclay. The EXULITE consortium is developing an LPP source that synchronizes 10 laser oscillators to achieve a repetition rate of ≥10 kHz and average power of 5 kW. The ELSAC prototype is complete, and manufacturing is underway.